Publikationsserver der Ostbayerischen Technischen Hochschule Regensburg
Not a member yet
6172 research outputs found
Sort by
Tyche algorithm: Markov models for generating learning paths in learning management systems
No full textIn the intricate tapestry of the cosmos, where celestial threads weave stories of destiny, an enigmatic deity stands at the intersection of chance and fate — Tyche, the goddess of fortune. In science and probability, coincidence plays a distinctive role in Bayesian Networks (BNs) and Markov Models (MMs). This paper introduces the Tyche algorithm named after the goddess of fortune. The Tyche algorithm is a Markov model designed to generate learning paths in Learning Management Systems (LMSs). A learning path is a type of individualization that personalises the order of learning elements within an LMS course. Learning elements are fundamental components within an LMS course, depicting the learning content in diverse ways. In our case, the learning path tailoring is based on the learner’s learning style according to Felder-Silverman Learning Style Model (FSLSM) – an indicator for the ideal pathway and learning element for the learner’s optimum learning.
The Tyche algorithm offers the advantage to provide students the most suitable learning path. Tyche is a MM structure with various matrices containing transition probabilities depending on the learning style. Nine categories of learning elements defined in a previous survey form the basis of the node structure of the MM. For a generic approach, a survey was designed to obtain the transition probabilities depending on the individual learning style. The survey with more than 100 German students participated is processed with the tool LimeSurvey. Students are asked about their learning style using the Index of Learning Styles (ILS) questionnaire according to Felder-Silverman and about the percentage probabilities of learning elements to get their individual sequence as learning path. The percentages are queried in two different ways. Firstly, the students were asked to provide a sequence of learning elements within each position of the sequence filled with probabilities for all learning elements. Secondly, a learning element is given and the students are asked to indicate which learning element they would work on next. The first way of asking for the probabilities is used to find the start node in the MM, whereas the second approach forms the probability matrices between the nodes within the MM.
As result of the survey, the Tyche algorithm presents generic transition probabilities. It improves the learning process of individuals only by asking for their learning style: it generates individual learning paths through the learning elements within an LMS based on the MM explained above by solely getting the answers of the ILS questionnaire as input. In the future, other questionnaires such as BFI-10 for personality traits or LIST-K for learning strategies may offer a more comprehensive input. However, the next step is to evaluate Tyche with about 25 students in a software engineering lecture. This is planned for the year 2024
Motion-Corrected Moving Average: Including Post-Hoc Temporal Information for Improved Video Segmentation
No full textReal-time computational speed and a high degree of precision are requirements for computer-assisted interventions. Applying a segmentation network to a medical video processing task can introduce significant inter-frame prediction noise. Existing approaches can reduce inconsistencies by including temporal information but often impose requirements on the architecture or dataset. This paper proposes a method to include temporal information in any segmentation model and, thus, a technique to improve video segmentation performance without alterations during training or additional labeling. With Motion-Corrected Moving Average, we refine the exponential moving average between the current and previous predictions. Using optical flow to estimate the movement between consecutive frames, we can shift the prior term in the moving-average calculation to align with the geometry of the current frame. The optical flow calculation does not require the output of the model and can therefore be performed in parallel, leading to no significant runtime penalty for our approach. We evaluate our approach on two publicly available segmentation datasets and two proprietary endoscopic datasets and show improvements over a baseline approach
Hard shell, soft core? Multi-disciplinary and multi-national insights into mental toughness among surgeons
No full textBackground:
With the prevalence of burnout among surgeons posing a significant threat to healthcare outcomes, the mental toughness of medical professionals has come to the fore. Mental toughness is pivotal for surgical performance and patient safety, yet research into its dynamics within a global and multi-specialty context remains scarce. This study aims to elucidate the factors contributing to mental toughness among surgeons and to understand how it correlates with surgical outcomes and personal well-being.
Methods:
Utilizing a cross-sectional design, this study surveyed 104 surgeons from English and German-speaking countries using the Mental Toughness Questionnaire (MTQ-18) along with additional queries about their surgical practice and general life satisfaction. Descriptive and inferential statistical analyses were applied to investigate the variations in mental toughness across different surgical domains and its correlation with professional and personal factors.
Results:
The study found a statistically significant higher level of mental toughness in micro-surgeons compared to macro-surgeons and a positive correlation between mental toughness and surgeons' intent to continue their careers. A strong association was also observed between general life satisfaction and mental toughness. No significant correlations were found between the application of psychological skills and mental toughness.
Conclusion:
Mental toughness varies significantly among surgeons from different specialties and is influenced by professional dedication and personal life satisfaction. These findings suggest the need for targeted interventions to foster mental toughness in the surgical community, potentially enhancing surgical performance and reducing burnout. Future research should continue to explore these correlations, with an emphasis on longitudinal data and the development of resilience-building programs
Artificial Intelligence (AI) improves endoscopists’ vessel detection during endoscopic submucosal dissection (ESD)
No full textAims
While AI has been successfully implemented in detecting and characterizing colonic polyps, its role in therapeutic endoscopy remains to be elucidated. Especially third space endoscopy procedures like ESD and peroral endoscopic myotomy (POEM) pose a technical challenge and the risk of operator-dependent complications like intraprocedural bleeding and perforation. Therefore, we aimed at developing an AI-algorithm for intraprocedural real time vessel detection during ESD and POEM.
Methods
A training dataset consisting of 5470 annotated still images from 59 full-length videos (47 ESD, 12 POEM) and 179681 unlabeled images was used to train a DeepLabV3+neural network with the ECMT semi-supervised learning method. Evaluation for vessel detection rate (VDR) and time (VDT) of 19 endoscopists with and without AI-support was performed using a testing dataset of 101 standardized video clips with 200 predefined blood vessels. Endoscopists were stratified into trainees and experts in third space endoscopy.
Results
The AI algorithm had a mean VDR of 93.5% and a median VDT of 0.32 seconds. AI support was associated with a statistically significant increase in VDR from 54.9% to 73.0% and from 59.0% to 74.1% for trainees and experts, respectively. VDT significantly decreased from 7.21 sec to 5.09 sec for trainees and from 6.10 sec to 5.38 sec for experts in the AI-support group. False positive (FP) readings occurred in 4.5% of frames. FP structures were detected significantly shorter than true positives (0.71 sec vs. 5.99 sec).
Conclusions
AI improved VDR and VDT of trainees and experts in third space endoscopy and may reduce performance variability during training. Further research is needed to evaluate the clinical impact of this new technology
Schärfeurteile für tonhaltige Geräusche und die Konsequenzen für die instrumentelle psychoakustische Analyse
No full textIn einer vorrangegangenen Studie wurde der Einfluss psychoakustischer Parameter auf Präferenzurteile für konsonante und dissonante Geräusche untersucht. Diese Geräusche hatten die gleiche Lautheit und basierten auf einem festen Rauschhintergrund und eingebetteten hochfrequenten Tonkomponenten. Es zeigte sich, dass die Präferenzurteile eines Probandenkollektives durch ebenfalls erhobene Subjektivurteile zu Schärfe, Tonhaltigkeit und Dissonanz mit hoher Präzision vorhergesagt werden konnten.Durch das angewandte Versuchsdesign konnte diese rein auf Subjektivbeurteilungen bestehende Vorhersage mit einer instrumentellen Vorhersage der Präferenzurteile verglichen werden. Der Vergleich ergab einen niedrigeren Korrelationskoeffizienten für die modellbasierte Vorhersage, was auch durch die sehr niedrige Korrelation (R2<0,2) der berechneten Schärfe nach DIN45692 mit den Schärfe-Subjektivurteilen zu begründen war. Eine mögliche Erklärung der unzureichenden Prognose der Schärfeurteile könnte in der Bewertungsstrategie der Proband:innen zu finden sein. Für die getesteten Geräusche mit auffälligen tonalen Komponenten könnte statt der Schärfe des Gesamtgeräuschs nur die Schärfe des tonalen Geräuschobjekts berücksichtigt worden sein. Die Schärfe-Analyse der auf ihren tonalen Anteil reduzierten Signale zeigt eine sehr gute Übereinstimmung mit den Subjektivurteilen, was die Hypothese einer Trennung von Ton- und Rauschobjekten unterstützt. Es wird diskutiert, ob und wie das Paradigma der akustischen Objekttrennung in der Modellierung psychoakustischer Empfindungsgrößen Anwendung finden kann
Report of RILEM TC 281-CCC: Insights into factors affecting the carbonation rate of concrete with SCMs revealed from data mining and machine learning approaches
No full textThe RILEM TC 281–CCC ‘‘Carbonation of concrete with supplementary cementitious materials’’ conducted a study on the effects of supplementary cementitious materials (SCMs) on the carbonation rate of blended cement concretes and mortars. In this context, a comprehensive database has been established, consisting of 1044 concrete and mortar mixes with their associated carbonation depth data over time. The dataset comprises mix designs with a large variety of binders with up to 94% SCMs, collected from the literature as well as unpublished testing reports. The data includes chemical composition and physical properties of the raw materials, mix-designs, compressive strengths, curing and carbonation testing conditions. Natural carbonation was recorded for several years in many cases with both indoor and outdoor results. The database has been analysed to investigate the effects of binder composition and mix design, curing and preconditioning, and relative humidity on the carbonation rate. Furthermore, the accuracy of accelerated carbonation testing as well as possible correlations between compressive strength and carbonation resistance were evaluated. The analysis revealed that the w/CaOreactive ratio is a decisive factor for carbonation resistance, while curing and exposure conditions also influence carbonation. Under natural exposure conditions, the carbonation data exhibit significant variations. Nevertheless, probabilistic inference suggests that both accelerated and natural carbonation processes follow a square-root-of-time behavior, though accelerated and natural carbonation cannot be converted into each other without corrections. Additionally, a machine learning technique was employed to assess the influence of parameters governing the carbonation progress in concretes
Data-Proximal Complementary ℓ1-TV Reconstruction for Limited Data Computed Tomography
No full textIn a number of tomographic applications, data cannot be fully acquired, resulting in severely underdetermined image reconstruction. Conventional methods in such cases lead to reconstructions with significant artifacts. To overcome these artifacts, regularization methods are applied that incorporate additional information. An important example is TV reconstruction, which is known to be efficient in compensating for missing data and reducing reconstruction artifacts. On the other hand, tomographic data are also contaminated by noise, which poses an additional challenge. The use of a single regularizer must therefore account for both the missing data and the noise. A particular regularizer may not be ideal for both tasks. For example, the TV regularizer is a poor choice for noise reduction over multiple scales, in which case ℓ1 curvelet regularization methods are well suited. To address this issue, in this paper, we present a novel variational regularization framework that combines the advantages of different regularizers. The basic idea of our framework is to perform reconstruction in two stages. The first stage is mainly aimed at accurate reconstruction in the presence of noise, and the second stage is aimed at artifact reduction. Both reconstruction stages are connected by a data proximity condition. The proposed method is implemented and tested for limited-view CT using a combined curvelet–TV approach. We define and implement a curvelet transform adapted to the limited-view problem and illustrate the advantages of our approach in numerical experiments
Assessment of implant internal stresses under physiological femoral loading: Translation to a simplified bending load model
No full textThe success of surgical treatment for fractures hinges on various factors, notably accurate surgical indication. The process of developing and certifying a new osteosynthesis device is a lengthy and costly process that requires multiple cycles of review and validation. Current methods, however, often rely on predecessor standards rather than physiological loads in specific anatomical locations. This study aimed to determine actual loads experienced by an osteosynthesis plate, exemplified by a standard locking plate for the femoral shaft, utilizing finite elements analysis (FEA) and to obtain the bending moments for implant development standard tests. A protocol was developed, involving the creation and validation of a fractured femur model fixed with a locking plate, mechanical testing, and FEA. The model’s validation demonstrated exceptional accuracy in predicting deformations, and the FEA revealed peak stresses in the fracture bridging zone. Results of a parametric analysis indicate that larger fracture gaps significantly impact implant mechanical behavior, potentially compromising stability. This study underscores the critical need for realistic physiological conditions in implant evaluations, providing an innovative translational approach to identify internal loads and optimize implant designs. In conclusion, this research contributes to enhancing the understanding of implant performance under physiological conditions, promoting improved designs and evaluations in fracture treatments
The ADVanced Organ Support (ADVOS) hemodialysis system removes IL-6: an in vitro proof-of-concept study
No full textBACKGROUND
IL-6 is a pleiotropic cytokine modulating inflammation and metabolic pathways. Its proinflammatory effect plays a significant role in organ failure pathogenesis, commonly elevated in systemic inflammatory conditions. Extracorporeal blood purification devices, such as the Advanced Organ Support (ADVOS) multi hemodialysis system, might offer potential in mitigating IL-6's detrimental effects, yet its efficacy remains unreported.
METHODS
We conducted a proof-of-concept in vitro study to assess the ADVOS multi system's efficacy in eliminating IL-6. Varying concentrations of IL-6 were introduced into a swine blood model and treated with ADVOS multi for up to 12 h, employing different blood and concentrate flow rates. IL-6 reduction rate, clearance, and dynamics in blood and dialysate were analyzed.
RESULTS
IL-6 clearance rates of 0.70 L/h and 0.42 L/h were observed in 4 and 12-h experiments, respectively. No significant differences were noted across different initial concentrations. Reduction rates ranged between 40 and 46% within the first 4 h, increasing up to 72% over 12 h, with minimal impact from flow rate variations. Our findings suggest that an IL-6-albumin interaction and convective filtration are implicated in in vitro IL-6 elimination with ADVOS multi.
CONCLUSIONS
This study demonstrates for the first time an efficient and continuous in vitro removal of IL-6 by ADVOS multi at low blood flow rates. Initial concentration-dependent removal transitions to more consistent elimination over time. Further clinical investigations are imperative for comprehensive data acquisition